the Ovule of Myrica Gale. 357 
a section a little higher in the series shows the edges. This is shown 
in longitudinal section in Fig. 2. 
Generally in the ovules of recent plants, the vascular supply ter- 
minates here, i. e. in a flat plate of tracheids at the base of the nucellus, 
but in Myrica Gale , the edges of this cup of tracheids run out into eight 
or nine slender vascular strands, each composed of a few tracheids and 
a small quantity of phloem, which traverse the single integument (Text- 
fig. I, 7 i, and Fig. 2 i.b.). 
A tangential section of the integument showing these vascular strands 
is represented in Fig. 3. The strands pass up the integument without 
branching, getting smaller as they get higher, until, almost at the apex 
of the ovule, the strand which has ’diminished to a single tracheid dies 
out (Fig. 2 i.b.). 
In a transverse section of the ovule, therefore, one finds a ring of 
eight to nine small vascular bundles as seen in Text-fig. I, 7 i. 
Running up the centre of the nucellus from its base to the base of the 
embryo-sac is a strand of tissue, consisting of parenchymatous cells con- 
siderably elongated in the vertical direction (Fig. 2 c.s.). This strand 
of tissue probably facilitates the conduction of food material to the 
embryo, since it connects the cells at the base of the nucellus which 
have very dense contents, with the embryo-sac. 
These characters, i. e. the free nucellus and the vascular supply running 
up into the integument are remarkable — occurring as they do in a family 
of recent plants, and may be regarded as being of phylogenetic interest. 
In living Angiosperms the nucellus and integument are generally 
fused together almost to the apex, and the integument is devoid of 
vascular tissue. 
It will be remembered, however, that a free nucellus and vascular 
nteguments were prevalent in many of the fossil plants. 
The ovules of some of the Pteridosperms and many of the fossil 
Gymnosperms have been described as having a free nucellus. Almost 
all the fossil seeds described by Brongniart 1 in his ‘ Recherches sur les 
graines fossiles silicifiees ’ are represented as having a nucellus quite free 
from the surrounding integument. 
Trigonocarpus , probably the seed of Medullosa, , recently described 
by Scott and Maslen, 2 and the ovule of Myrica Gale show a great 
similarity of structure (Text-fig. II). Both of them possess a single 
integument, and in each case the nucellus stands up freely within it. 
This character of a free nucellus in these older fossil seeds may indicate 
that the integument had not as yet become an integral part of the seed. 
1 Brongniart (’ 81 ) : Recherches sur les graines fossiles silicifiees. Paris. 
2 Scott and Maslen: The Structure of the Palaeozoic Seeds, Trigonocarpus Parkinsoni , 
Brongniart, and Trig. Oliveri , sp. nov. Ann. of Bot, vol. xxi, Jan., 1907. 
